Thickness detecting device for flat panel display substrate

Abstract

The application discloses a kind of thickness detection devices for flat panel display substrate, it is related to substrate detection field, including test bearing platform, the right side of the test bearing platform is equipped with path correction unit, the inside symmetry of the test bearing platform is equipped with two laser testing components, the middle of the test bearing platform is provided with motion support component, the top end symmetry of the motion support component is equipped with light compensation component, the side of the top end of the motion support component is equipped with the motion monitoring unit of providing jitter compensation, the other side of the top end of the motion support component is equipped with displacement adjustment module;By introducing the double dynamic compensation mechanism of motion compensation and light compensation, the main interference source in the measurement process is eliminated, so that the thickness data finally obtained is closer to the real state of the substrate, greatly reduces the false alarm and the missing report rate, not only can timely find defects, but also can help users quickly locate the root cause of process problem, to guide the improvement of production parameters.

Description

Technical Field

[0001] This invention relates to the field of substrate inspection technology, specifically to a thickness inspection device for flat panel display substrates. Background Technology

[0002] With the rapid development of the flat panel display industry, high-precision display technologies such as OLED, Mini-LED, and Micro-LED have become mainstream. The glass substrates, silicon substrates, and flexible films involved in these technologies are becoming increasingly thinner and larger, requiring uniformity in thickness down to the micrometer or even nanometer level. Any minute thickness unevenness can lead to fatal defects such as uneven light emission, color deviation, and short circuits. Thickness is one of the core indicators for measuring substrate quality. During the manufacturing process, 100% online or offline thickness inspection is crucial for ensuring product yield and reducing production costs. Traditional sampling inspection methods can no longer meet the demands of high-quality control.

[0003] Traditional laser thickness gauges are highly susceptible to changes in ambient light, substrate surface roughness, material properties, and refraction and scattering caused by surface films, leading to significant noise and errors in measurement data. This necessitates frequent manual re-inspections. In high-speed, wide-range scanning measurements, vibrations of the motion platform, lead screw backlash, and start-stop jitter are unavoidable. Minor motion jitters directly superimpose onto the laser displacement sensor readings, distorting the thickness data. Many existing systems can only determine if an anomaly is present but cannot provide inspectors with specific data on the shape, area, and trend of the anomaly. They lack in-depth analysis of defect morphology and cannot provide valuable information for process improvement. After an anomaly is detected, it is usually necessary to stop the machine or manually reset the detection path for re-inspection and confirmation, disrupting the production rhythm and resulting in low efficiency. Summary of the Invention

[0004] (a) Technical problems to be solved

[0005] In view of the above-mentioned shortcomings of the prior art, the present invention provides a thickness detection device for flat panel display substrates, which can effectively solve the problems of the prior art.

[0006] (II) Technical Solution

[0007] To achieve the above objectives, the present invention provides the following technical solution:

[0008] This invention discloses a thickness detection device for flat panel display substrates, including a test platform. A path correction unit is installed on the right side of the test platform. Two laser testing components are symmetrically installed inside the test platform. A motion support component is disposed in the middle of the test platform. Light compensation components are symmetrically installed on the top of the motion support component. A motion monitoring unit providing jitter compensation is installed on one side of the top of the motion support component. A displacement adjustment module is installed on the other side of the top of the motion support component. A light source monitoring module providing light source interference compensation is installed inside the test platform. Wherein:

[0009] The test platform serves as the main support for the system's functional components.

[0010] The laser testing component is used to perform laser scanning detection on the substrate thickness and output test data.

[0011] The displacement adjustment module is used to adjust and edit the control commands for the motion trajectory of the motion support components;

[0012] The motion support component is used to receive control commands from the displacement adjustment module and transport the substrate under test along a predetermined motion trajectory to the test area of ​​the laser testing component.

[0013] The path correction unit is used to receive test data, determine whether the thickness is abnormal, and if there is an abnormality, it calls the region recognition model built based on the strong learning algorithm to identify the shape of the abnormal region, extract key shape features to mark abnormal points, and generate the optimized displacement path of the motion support component based on all abnormal point marks.

[0014] Furthermore, the path correction unit is further equipped with sub-modules, including a data reading module, an anomaly detection module, an area identification module, a point marking module, and a trajectory analysis module. The anomaly detection module is interconnected with the data reading module and the area identification module via a wireless network, and the point marking module is interconnected with the area identification module and the trajectory analysis module via a wireless network.

[0015] The data reading module is used to integrate and temporarily store the test data after motion compensation by the motion monitoring unit and light compensation by the light compensation component;

[0016] The anomaly detection module is used to compare the test data obtained by the data reading module with the standard substrate parameters to determine the thickness anomaly. If an anomaly is detected, the anomaly data is submitted to the area identification module.

[0017] The region identification module is triggered when the anomaly judgment module determines that there is an anomaly. It uses a strong learning algorithm to build a region identification model, inputs thickness anomaly data, and outputs shape data of the current substrate anomaly region, including boundary contour, area and geometric features.

[0018] The point marking module is used to extract the shape features of abnormal areas, extract several shape features, and screen key features whose influence coefficient exceeds the threshold to mark points in the virtual coordinate system of the substrate.

[0019] The trajectory analysis module is used to generate the displacement path of the displacement adjustment module within a preset period based on the abnormal point markers.

[0020] Furthermore, after receiving the thickness anomaly data transmitted by the anomaly judgment module, the region recognition module maps the anomaly data into a two-dimensional thickness distribution matrix according to the spatial position of the substrate; it extracts features from the two-dimensional thickness distribution matrix using a pre-trained region recognition model; it inputs the extracted feature vectors into a fully connected layer to predict the boundary of the anomaly region, and outputs a coordinate sequence containing the contour vertices of the anomaly region; it calculates the corresponding geometric attributes of the anomaly region based on the coordinate sequence; and it combines the contour vertex coordinate sequence with the geometric attributes to form shape data and outputs it to the point marking module.

[0021] Furthermore, when generating the displacement path, the trajectory analysis module performs the following operations: dividing the marked and unmarked areas in the virtual coordinate system based on the abnormal point markers; calculating the shortest detection path that covers all unmarked abnormal areas while avoiding the marked normal areas; and discretizing the calculated optimal path into an ordered sequence of coordinate points, which serves as the trajectory command for the displacement adjustment module to control the movement of the motion support component in the next preset cycle.

[0022] Furthermore, the light source monitoring module is used to detect the intensity, incident angle and frequency of change of ambient light, identify environmental interference factors in real time, obtain light source identification data, and determine whether light source compensation is needed based on preset light source evaluation standards. If the determination is yes, the light compensation component is triggered.

[0023] The optical compensation components are symmetrically and evenly distributed around the laser testing components and move synchronously with the motion support components. They are used to receive light source identification data provided by the light source monitoring module and dynamically adjust the intensity of their own compensation light source based on the intensity, direction, and spectrum information of the interfering light source in the light source identification data to provide targeted compensation for optical interference factors. When the light source identification data indicates that the current optical interference exceeds the maximum compensation capability range of the optical compensation components, the optical compensation components output an alarm signal.

[0024] Furthermore, the environmental interference factors include: the refractive and scattering characteristics of the current substrate surface itself, and ambient light.

[0025] Furthermore, the motion support component includes an electric telescopic rod one, an electric telescopic rod two, and a support frame. The bottom end of the outer rod of the electric telescopic rod one is fixedly connected to the top end of the test platform. One end of the inner rod of the electric telescopic rod one is fixedly connected to the surface of the outer rod of the electric telescopic rod two. One end of the inner rod of the electric telescopic rod two is fixedly connected to the surface of the support frame. The electric telescopic rod one and the electric telescopic rod two adjust their telescopic process according to the control command of the displacement adjustment module to control the position of the base plate on the support frame.

[0026] Furthermore, the light source evaluation standard is set based on a preset signal-to-noise ratio threshold and the working spectrum range of the laser test component. When the energy ratio of the monitored ambient light in the working spectrum range of the laser exceeds a preset ratio or the energy dispersion of the laser spot caused by scattering on the substrate surface exceeds a preset dispersion threshold, it is determined that light source compensation is required.

[0027] Furthermore, after receiving the displacement path provided by the path correction unit, the displacement adjustment module parses the displacement path and converts it into a sequence of displacement adjustment instructions for the motion support component in the horizontal direction. It controls the motion support component to move according to the instruction sequence. During this process, the laser testing component acquires the adjusted test data and submits the adjusted test data back to the path correction unit for retesting and re-judgment of the original abnormal area. If the abnormality is still determined after retesting, the area recognition model of the path correction unit is used to further identify and plan the shape data of the current abnormal area of ​​the substrate. The final planned abnormal area shape data is then submitted and displayed through the path correction unit.

[0028] Furthermore, the motion monitoring unit is further equipped with sub-modules, including: a motion acquisition module, a motion compensation module, and a data application module, wherein:

[0029] The motion acquisition module is used to obtain the jitter coefficient during the current operation of the device through vibration sensors;

[0030] The motion compensation module is used to analyze the current jitter coefficient to determine whether it exceeds a preset threshold. If it does, it calculates compensation for the test data obtained by the laser test component in the current cycle based on the extent of the jitter coefficient exceeding the threshold.

[0031] The data application module is used to fuse the compensation calculation results obtained from the motion compensation module with the current test data.

[0032] Furthermore, the formula for the motion compensation module to perform compensation calculations is as follows:

[0033] ;

[0034] In the formula, This represents the thickness data after compensation. Represents the original thickness measurement value. This represents the jitter acceleration value in the Z-axis direction. Represents the sampling time interval. This represents the velocity at the end of the Z-axis in the previous sampling period, with an initial value of 0. This represents the equipment rigidity coefficient.

[0035] (III) Beneficial Effects

[0036] Compared with the known prior art, the technical solution provided by this invention has the following beneficial effects:

[0037] 1. By introducing a dual dynamic compensation mechanism of motion compensation and optical compensation, the main sources of interference in the measurement process are eliminated. Motion compensation directly offsets the measurement errors caused by the vibration and jitter of the equipment itself, ensuring the stability of data acquisition. Optical compensation effectively suppresses the interference of ambient light changes and refraction and scattering on the substrate surface on laser measurement, ensuring the authenticity of the readings. This makes the final thickness data closer to the true state of the substrate, greatly reducing the false alarm and false alarm rates, and providing a reliable data foundation for high-quality production.

[0038] 2. Through the path correction unit, strong learning algorithms and feature extraction technology are used to achieve deep understanding of defects. The region recognition module can automatically identify the shape, size and location of abnormal regions, transforming disordered abnormal data into meaningful morphological information. The point marking module can further filter out key feature points and accurately locate the core of the problem. This enables the system to not only discover defects in a timely manner, but also help users quickly locate the root cause of process problems, thereby guiding the improvement of production parameters and improving the yield rate.

[0039] 3. Through the trajectory analysis module, after an anomaly is detected, the system automatically generates the optimal retest path and drives the actuator to perform precise retesting. No manual intervention is required. The entire process is completed autonomously by the system, avoiding production interruptions and time delays caused by manual shutdown, manual positioning, and remeasurement in traditional methods. This greatly improves the pace and efficiency of online inspection, adapts to the rhythm of high-speed production lines, and reduces reliance on manpower. Attached Figure Description

[0040] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the accompanying drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are merely some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without any creative effort.

[0041] Figure 1 This is a schematic diagram of the overall three-dimensional structure of the present invention;

[0042] Figure 2 This is a schematic diagram of the overall three-dimensional structure of the present invention from another angle;

[0043] Figure 3 This is a three-dimensional structural diagram of the motion support component in this invention;

[0044] Figure 4 This is a schematic diagram of the overall framework of the present invention;

[0045] Figure 5 This is a schematic diagram of the path correction unit in this invention;

[0046] Figure 6 This is a schematic diagram of the motion monitoring unit in this invention.

[0047] The labels in the diagram represent: 1. Test platform; 2. Laser testing component; 3. Displacement adjustment module; 4. Motion support component; 41. Electric telescopic pole one; 42. Electric telescopic pole two; 43. Support frame; 5. Path correction unit; 51. Data reading module; 52. Anomaly detection module; 53. Area recognition module; 54. Point marking module; 55. Trajectory analysis module; 6. Light source monitoring module; 7. Light compensation component; 8. Motion monitoring unit; 81. Motion acquisition module; 82. Motion compensation module; 83. Data application module. Detailed Implementation

[0048] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention.

[0049] The present invention will be further described below with reference to embodiments.

[0050] This embodiment provides a thickness detection device for a flat panel display substrate, such as... Figures 1-6As shown, the test platform includes a test stand 1. A path correction unit 5 is installed on the right side of the test stand 1. Two laser testing components 2 are symmetrically installed inside the test stand 1. A motion support component 4 is located in the middle of the test stand 1. Light compensation components 7 are symmetrically installed on the top of the motion support component 4. A motion monitoring unit 8 providing jitter compensation is installed on one side of the top of the motion support component 4. A displacement adjustment module 3 is installed on the other side of the top of the motion support component 4. A light source monitoring module 6 providing light source interference compensation is installed inside the test stand 1. Wherein:

[0051] Test platform 1 serves as the main support for the system's functional components.

[0052] Laser testing component 2 is used to perform laser scanning detection on the substrate thickness, output test data, and place the substrate between two laser testing components 2.

[0053] The displacement adjustment module 3 is used to adjust and edit the control commands for the motion trajectory of the motion support component 4. After receiving the displacement path provided by the path correction unit 5, the displacement adjustment module 3 parses the displacement path and converts it into a sequence of horizontal displacement adjustment commands for the motion support component 4. It controls the motion support component 4 to move according to the command sequence. During this process, the laser testing component 2 acquires the adjusted test data and re-submits the adjusted test data to the path correction unit 5 for retesting and re-judgment of the original abnormal area. If it is still determined to be abnormal after retesting, the area recognition model of the path correction unit 5 is used to further identify and plan the shape data of the current abnormal area of ​​the substrate. The final planned abnormal area shape data is submitted and displayed through the path correction unit 5. The dynamic adjustment and editing of the motion trajectory of the motion support component 4, by receiving the displacement path provided by the path correction unit 5 and parsing it into a command sequence, achieves precise positioning of the substrate, avoids invalid scanning, and reduces time waste. If it is still abnormal after retesting, the model of the path correction unit 5 is called to further identify the abnormal shape, enhancing the system's adaptability to complex defects, optimizing the detection process, and improving the accuracy of abnormal area identification.

[0054] The motion support component 4 is used to receive control commands from the displacement adjustment module 3 and transport the substrate to be tested along a predetermined motion trajectory to the test area of ​​the laser test component 2. The motion support component 4 includes an electric telescopic rod 41, an electric telescopic rod 42, and a support frame 43. The bottom end of the outer rod of the electric telescopic rod 41 is fixedly connected to the top end of the test support platform 1. One end of the inner rod of the electric telescopic rod 41 is fixedly connected to the surface of the outer rod of the electric telescopic rod 42. One end of the inner rod of the electric telescopic rod 42 is fixedly connected to the surface of the support frame 43. The electric telescopic rods 41 and 42 adjust their telescopic process according to the control commands from the displacement adjustment module 3 to control the position of the substrate on the support frame 43.

[0055] The path correction unit 5 is used to receive test data and determine whether the thickness is abnormal. If there is an abnormality, it calls the region recognition model built based on the strong learning algorithm to identify the shape of the abnormal region and extract key shape features to mark the abnormal points. Based on all the abnormal point marks, it generates the optimized displacement path of the motion support component 4.

[0056] The path correction unit 5 has sub-modules deployed below it. These sub-modules include a data reading module 51, an anomaly detection module 52, an area identification module 53, a point marking module 54, and a trajectory analysis module 55. The anomaly detection module 52 is interconnected with the data reading module 51 and the area identification module 53 via a wireless network. The point marking module 54 is interconnected with the area identification module 53 and the trajectory analysis module 55 via a wireless network.

[0057] Data reading module 51 is used to integrate and temporarily store the test data after motion compensation by motion monitoring unit 8 and light compensation by light compensation component 7;

[0058] The anomaly detection module 52 is used to compare the test data obtained by the data reading module 51 with the standard substrate parameters to determine the thickness anomaly. If an anomaly is detected, the anomaly data is submitted to the area identification module 53.

[0059] The region identification module 53 is triggered when the anomaly judgment module 52 determines that an anomaly exists. It employs a strong learning algorithm to construct a region identification model, inputs thickness anomaly data, and outputs shape data of the preliminarily identified current substrate anomaly region, including boundary contours, area, and geometric features. After receiving the thickness anomaly data from the anomaly judgment module 52, the region identification module 53 maps the anomaly data to a two-dimensional thickness distribution matrix according to the substrate's spatial location. It then extracts features from the two-dimensional thickness distribution matrix using a pre-trained region identification model. The feature extraction process involves alternating three convolutional operations and two pooling operations, where convolutional operations identify local thickness abrupt change patterns, and pooling operations compress spatial dimensions while retaining significant features. The extracted feature vectors are input into a fully connected layer for anomaly region boundary prediction, outputting a coordinate sequence containing the vertices of the anomaly region contour. Based on the coordinate sequence, the corresponding geometric attributes of the anomaly region are calculated, including at least the contour type, area value, maximum length dimension, and center point coordinates. Finally, the contour vertex coordinate sequence and geometric attributes are combined into shape data and output to the point marking module 54.

[0060] The point marking module 54 is used to extract the shape features of the abnormal area, including area, perimeter, roundness, rectangularity, centroid position, etc., and to screen key features with influence coefficients exceeding the threshold. Point marking is performed in the virtual coordinate system of the substrate. The abnormal point marking includes marking the centroid, boundary inflection points and feature extreme points of the abnormal area, and generating a marking file containing the coordinates of these points and the abnormal level.

[0061] The trajectory analysis module 55 is used to generate the displacement path of the displacement adjustment module 3 within a preset period based on the abnormal point markings. When generating the displacement path, the trajectory analysis module 55 performs the following operations: dividing the marked area and the unmarked area in the virtual coordinate system according to the abnormal point markings; calculating the shortest detection path that covers all unmarked abnormal areas while avoiding the marked normal areas; discretizing the calculated optimal path into an ordered coordinate point sequence, which serves as the trajectory command for the displacement adjustment module 3 to control the movement of the motion support component 4 in the next preset period, so that the motion support component 4 can perform directional displacement of the substrate under test according to the optimized path.

[0062] This embodiment improves the accuracy and efficiency of substrate thickness detection by integrating laser scanning, light source compensation, jitter compensation, and an intelligent anomaly recognition system based on strong learning algorithms. Compared with existing technologies, this device can adjust the detection path in real time, automatically identify and mark abnormal areas, and achieve precise positioning by optimizing the displacement path. This reduces manual intervention and errors, improves the automation level and reliability of detection, and greatly enhances production efficiency.

[0063] At other levels, this embodiment provides an optimization mechanism for motion compensation and light compensation. The light source monitoring module 6 is used to detect the intensity, incident angle and frequency of change of ambient light, identify environmental interference factors in real time, obtain light source identification data, and determine whether light source compensation is needed based on preset light source evaluation standards. If the determination is yes, the light compensation component 7 is triggered. Environmental interference factors include: the refractive characteristics and scattering characteristics of the current substrate surface and ambient light.

[0064] The optical compensation component 7 is symmetrically and uniformly distributed around the laser testing component 2 and moves synchronously with the motion support component 4. It is used to receive the light source identification data provided by the light source monitoring module 6 and dynamically adjust the intensity of its own compensation light source according to the intensity, direction and spectrum information of the interfering light source in the light source identification data to provide targeted compensation for optical interference factors. When the light source identification data indicates that the current optical interference exceeds the maximum compensation capability range of the optical compensation component 7, the optical compensation component 7 outputs an alarm signal. The light source evaluation standard is set based on the preset signal-to-noise ratio threshold and the working spectrum range of the laser testing component 2. When the energy ratio of the monitored ambient light in the laser working spectrum range exceeds the preset ratio or the energy dispersion of the laser spot caused by the scattering of the substrate surface exceeds the preset dispersion threshold, it is determined that light source compensation is required.

[0065] Compared with existing technologies, this technology ensures the data stability of the laser testing component 2 in complex lighting environments, reduces measurement noise caused by light fluctuations, and improves the signal-to-noise ratio. It is especially suitable for semi-transparent or highly reflective display substrates, reduces optical errors caused by substrate material or surface unevenness, and enhances the reliability and repeatability of test data.

[0066] In this embodiment, the motion monitoring unit 8 has sub-modules deployed below it, including: a motion acquisition module 81, a motion compensation module 82, and a data application module 83, wherein:

[0067] The motion acquisition module 81 is used to acquire the jitter coefficient during the current operation of the device through a vibration sensor;

[0068] The motion compensation module 82 is used to analyze the current jitter coefficient and determine whether it exceeds the preset threshold. If it does, the module calculates compensation based on the magnitude of the jitter coefficient exceeding the threshold.

[0069] The data application module 83 is used to fuse the compensation calculation results obtained by the motion compensation module 82 with the current test data.

[0070] The formula used by motion compensation module 82 for compensation calculation is as follows:

[0071] ;

[0072] In the formula, This represents the thickness data after compensation. Represents the original thickness measurement value. This represents the jitter acceleration value in the Z-axis direction. Represents the sampling time interval. This represents the velocity at the end of the Z-axis in the previous sampling period, with an initial value of 0. This represents the equipment rigidity coefficient, with a value ranging from 0.8 to 1.2; when >0, when the platform accelerates upwards, the compensation amount is negative because the measured value is too large; <0, when the platform accelerates downwards, the compensation amount is positive because the measured value is too small.

[0073] Working principle: In the specific implementation of this invention, the test support platform 1 serves as the main support for each functional component and module, the laser test component 2 is installed on the test support platform 1, the motion support component 4 transports the substrate to be tested to the test area of ​​the laser test component according to a predetermined motion trajectory, and the displacement adjustment module 3 adjusts and controls the editing instructions of the motion trajectory of the motion displacement component.

[0074] The motion acquisition module 81 detects the motion state of the test platform 1 and the motion support component 4 during the substrate testing process, identifies motion jitter, and submits the jitter data exceeding the preset threshold to the motion compensation module 82. The motion compensation module 82 performs jitter compensation analysis based on the abnormal jitter data and adjusts the motion compensation of the test data obtained in the current cycle submitted by the laser test component 2 to the test platform 1.

[0075] The light source monitoring module 6 identifies the refraction, scattering, or ambient light on the substrate surface in the current monitoring environment, obtains light source identification data, and triggers the light compensation component 7. The light compensation component 7 is distributed around the laser test component 2. Based on the light source identification data provided by the current light source monitoring module 6, the light compensation component 7 provides light source compensation of corresponding intensity or direction. If the light compensation component 7 cannot complete the compensation, an alarm prompt will be issued.

[0076] The data reading module 51 acquires the final test data submitted by the basic test support platform 1 after light and motion compensation. The anomaly judgment module 52 analyzes the data obtained by the data reading module 51, compares it with the standard substrate parameters, and judges whether there is an anomaly. If there is an anomaly, the abnormal data is submitted to the area identification module 53.

[0077] The region identification module 53 constructs a region identification model using a strong learning algorithm. The region identification model reads abnormal data of the current substrate thickness unevenness parameter and outputs the shape data of the current substrate abnormal region that is initially identified. The point marking module 54 extracts several shape features based on the shape data of the current substrate thickness abnormal region, analyzes the influence coefficient of key shape features, filters key shape features with influence coefficients higher than a preset threshold, and marks abnormal points. The trajectory analysis module 55 generates the displacement path of the motion support component 4 within a preset period based on the obtained abnormal point markings. The displacement adjustment module 3 converts the displacement path into a displacement adjustment command for the motion support component 4 at a horizontal or tilt angle. The laser testing component 2 then obtains the adjusted test data, which is then handed over to the abnormal judgment module 52 for re-judgment of the abnormal region. The region identification model further identifies the shape data of the current substrate abnormal region and submits it to the path correction unit 5 for display.

[0078] The above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions will not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A thickness detection device for a flat panel display substrate, characterized in that, The test platform (1) includes a path correction unit (5) installed on the right side of the test platform (1), two laser test components (2) symmetrically installed inside the test platform (1), a motion support component (4) set in the middle of the test platform (1), a light compensation component (7) symmetrically installed on the top of the motion support component (4), a motion monitoring unit (8) providing jitter compensation installed on one side of the top of the motion support component (4), a displacement adjustment module (3) installed on the other side of the top of the motion support component (4), and a light source monitoring module (6) providing light source interference compensation installed inside the test platform (1). The test platform (1) serves as a system... The system is a carrier of functional components; a laser testing component (2) for laser scanning detection of substrate thickness and outputting test data; a displacement adjustment module (3) for adjusting and editing the control commands of the motion support component (4); a motion support component (4) for receiving control commands from the displacement adjustment module (3) and transporting the substrate to be tested along a predetermined motion trajectory to the test area of ​​the laser testing component (2); and a path correction unit (5) for receiving test data, determining whether the thickness is abnormal, and if there is an abnormality, calling a region recognition model based on a strong learning algorithm to identify the shape of the abnormal region, extracting key shape features to mark abnormal points, and generating a motion support based on all abnormal point marks. The optimized displacement path of component (4); the path correction unit (5) has a sub-module deployed below it, which includes a data reading module (51), an anomaly judgment module (52), an area identification module (53), a point marking module (54), and a trajectory analysis module (55). The anomaly judgment module (52) is connected to the data reading module (51) and the area identification module (53) via a wireless network. The point marking module (54) is connected to the area identification module (53) and the trajectory analysis module (55) via a wireless network. Among them: the data reading module (51) is used to integrate and temporarily store the test data after motion compensation by the motion monitoring unit (8) and light compensation by the light compensation component (7). According to; the anomaly judgment module (52) is used to compare the test data obtained by the data reading module (51) with the standard substrate parameters to determine the thickness anomaly. If the anomaly is determined, the anomaly data is submitted to the area recognition module (53); the area recognition module (53) is used to trigger when the anomaly judgment module (52) determines that there is an anomaly. It uses a strong learning algorithm to construct an area recognition model, inputs the thickness anomaly data, and outputs the shape data of the current substrate anomaly area, including the boundary contour, area and geometric features; the point marking module (54) is used to extract the shape features of the anomaly area, extract several shape features, and screen the key features with influence coefficients exceeding the threshold to mark the points in the virtual coordinate system of the substrate;When generating the displacement path, the trajectory analysis module (55) performs the following operations: dividing the marked and unmarked areas in the virtual coordinate system according to the abnormal point markers; calculating the shortest detection path covering all unmarked abnormal areas while avoiding the marked normal areas; discretizing the calculated optimal path into an ordered coordinate point sequence as the trajectory command for the displacement adjustment module (3) to control the movement of the motion support component (4) in the next preset cycle; the trajectory analysis module (55) is used to generate the displacement path of the displacement adjustment module (3) in the preset cycle according to the abnormal point markers; after receiving the thickness abnormal data transmitted by the abnormal judgment module (52), the region recognition module (53) maps the abnormal data into a two-dimensional thickness distribution matrix according to the spatial position of the substrate; extracts features from the two-dimensional thickness distribution matrix through a pre-trained region recognition model; inputs the extracted feature vectors into the fully connected layer to predict the boundary of the abnormal region and outputs a coordinate sequence containing the contour vertices of the abnormal region; calculates the geometric attributes corresponding to the abnormal region based on the coordinate sequence; and combines the contour vertex coordinate sequence with the geometric attributes into shape data and outputs it to the point marker module (54).

2. The thickness detection device for a flat panel display substrate according to claim 1, characterized in that, The motion support component (4) includes an electric telescopic rod one (41), an electric telescopic rod two (42), and a support frame (43). The bottom end of the outer rod of the electric telescopic rod one (41) is fixedly connected to the top end of the test support platform (1). One end of the inner rod of the electric telescopic rod one (41) is fixedly connected to the surface of the outer rod of the electric telescopic rod two (42). One end of the inner rod of the electric telescopic rod two (42) is fixedly connected to the surface of the support frame (43). The electric telescopic rod one (41) and the electric telescopic rod two (42) adjust the telescopic process according to the control command of the displacement adjustment module (3) to control the position of the base plate on the support frame (43).

3. The thickness detection device for a flat panel display substrate according to claim 1, characterized in that, The light source monitoring module (6) is used to detect the intensity, incident angle and frequency of ambient light, identify environmental interference factors in real time, obtain light source identification data, and determine whether light source compensation is required based on the preset light source evaluation standard. If the determination is yes, the light compensation component (7) is triggered. The environmental interference factors include: the refractive characteristics and scattering characteristics of the current substrate surface and ambient light. The light compensation component (7) is symmetrically and uniformly distributed around the laser test component (2) and moves synchronously with the motion support component (4). It is used to receive the light source identification data provided by the light source monitoring module (6) and dynamically adjust the intensity of its own compensation light source according to the intensity, direction and spectrum information of the interference light source in the light source identification data, so as to perform targeted compensation for optical interference factors. When the light source identification data indicates that the current optical interference exceeds the maximum compensation capability range of the light compensation component (7), the light compensation component (7) outputs an alarm signal.

4. The thickness detection device for a flat panel display substrate according to claim 3, characterized in that, The light source evaluation standard is based on the preset signal-to-noise ratio threshold and the working spectrum range of the laser test component (2). When the energy ratio of the monitored ambient light in the working spectrum range of the laser exceeds the preset ratio or the energy dispersion of the laser spot caused by the scattering of the substrate surface exceeds the preset dispersion threshold, it is determined that light source compensation is required.

5. The thickness detection device for a flat panel display substrate according to claim 1, characterized in that, After receiving the displacement path provided by the path correction unit (5), the displacement adjustment module (3) parses the displacement path and converts it into a sequence of displacement adjustment instructions for the motion support component (4) in the horizontal direction. It controls the motion support component (4) to act according to the instruction sequence. During this process, the laser test component (2) obtains the adjusted test data and re-submits the adjusted test data to the path correction unit (5) for retesting and re-judgment of the original abnormal area. If it is still determined to be abnormal after retesting, the area recognition model of the path correction unit (5) is used to further identify and plan the shape data of the current abnormal area of ​​the substrate. The final planned abnormal area shape data is submitted and displayed through the path correction unit (5).

6. The thickness detection device for a flat panel display substrate according to claim 1, characterized in that, The motion monitoring unit (8) has a sub-module deployed below it. The sub-module includes a motion acquisition module (81), a motion compensation module (82), and a data application module (83). The motion acquisition module (81) is used to acquire the jitter coefficient during the current operation of the device through a vibration sensor. The motion compensation module (82) is used to analyze the current jitter coefficient and determine whether it exceeds a preset threshold. If it does, the module will calculate the compensation for the test data obtained by the laser test component (2) in the current cycle based on the magnitude of the jitter coefficient exceeding the threshold. The data application module (83) is used to fuse the compensation calculation result obtained by the motion compensation module (82) with the current test data.

7. The thickness detection device for a flat panel display substrate according to claim 6, characterized in that, The formula for the motion compensation module (82) to perform compensation calculations is as follows: In the formula, This represents the thickness data after compensation. Represents the original thickness measurement value. This represents the jitter acceleration value in the Z-axis direction. Represents the sampling time interval. This represents the velocity at the end of the Z-axis in the previous sampling period, with an initial value of 0. Represents the equipment rigidity coefficient, when When the platform accelerates upwards, the compensation amount is negative because the measured value is too large; When the platform accelerates downwards, the compensation amount is positive because the measured value is too small.

Images